INDEX

Explanations

phrases indicating actions, particularly those that begin with 'to'

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ussion

-0.64

paralle

-0.63

strong

-0.62

spection

-0.62

å¦

-0.62

meta

-0.62

AAAAAAAA

-0.62

quickShipAvailable

-0.61

ational

-0.61

need

-0.60

POSITIVE LOGITS

 compensate

1.13

 acquire

1.11

 settle

1.06

 advertise

1.04

 reproduce

0.99

 participate

0.97

 maintain

0.96

 insure

0.93

 obtain

0.92

 educate

0.92

Activations Density 0.069%

phrases indicating actions, particularly those that begin with 'to'

No Comments

No Known Activations